« first day (21 days earlier)      last day (4400 days later) » 

12:18 AM
@All I believe I have hit upon a function that allows me to get the definition of an expression without triggering the full evaluation or digging around in *Values lists which is something that I've been wanting for a while now. The question and my own answer are here:
Q: How do I evaluate only one step of an expression?

SpartacusI am looking for a simple, robust way to evaluate an expression only one step, and return the result in a held form. The definition of a single step is ambiguous, and this itself is probably worthy of exploration. Some interpretations will raise the question of what should be returned. I am sp...

The solution seems almost too simple to be true. I would appreciate your testing it to find its limits, and perhaps expand them.
(it -> hit ... darn edit timeout)
@MrWizard Where does the HoldForm come from?
@MrWizard fixed
@MrWizard why does step@Module[{x = 1}, x + x] return 2? Is this what you expect? Yes, this is about the definition of a single step
or step[f[2*(1 + 1)]] returns f[4] while step[2*(1+1)] returns 2*2
1 hour later…
1:34 AM
hey @verbeia, Ted Ersek is here
have people thought about making CDF content embeddable in the site?
yes, but that's not for us to decide but for SE... it's too early to request anything of the sort though.
my hunch is that it'll be declined, because although it is nice to have the interactive features, it will slow down the experience considerably when waiting for the player to load
well then i guess that's a problem with using CDF anywhere
2:01 AM
I agree with yoda. MathJax is already a performance hit; a CDF player will make things a bit slower...
If someone went to the trouble of creating a CDF to embed, would it not probably be worth the wait?
I would wait, but a lot of other people tend to be the impatient sort...
2:30 AM
@AndrewMacFie the problem is that you don't know if it was worth waiting before you actually wait... I could support it if it were set up so that the page loads fast normally with just a graphics and there's a link saying "click to launch CDF version". That way, you need to wait only if you're interested
2:41 AM
23 days in (21 according to Area51) and I think we're doing just fine. We need to look at the FAQ question and the community ad proposal, again. But, we're doing very well, including completely filling out the top tiers of recommended high rep users! All in all, a good 3 weeks.
Yes, it has only been 3 weeks since we started beta.
For me, the scariest thing about these 3 weeks is that there are already questions on the site that I have never seen and only found out they existed by Related Questions list on another question.
2 hours later…
5:11 AM
looks like Heike is on track for a gold badge
Q: Plotting several functions

stoppleI'd like to plot a function of one real and one integer variable, but I don't want them all shown in the same 2-D plot - I'd like to see them as separate curves so I can see both 'axes', more like Plot3D works. I'm sure Mathematica can do this, but how? Edit: Rephrased, I'd like to Plot3D a fun...

5:53 AM
@yoda ...if Vitaliy gets a "Nice Answer" first.
@JM right... I'm guessing he will, and that Heike will bag one more upvote along the way
2 hours later…
8:01 AM
@yoda thanks, nice to know he is aloe. Interestingly he isn't registered, so doesn't show up in the user list. I didn't realise you could answer questions and gain rep while unregistered. I wonder how many other users who have posted are not registered.
@rcollyer thanks for the report, that is great news. I am still a bit surprised to see that I have more than 2000 rep already.
8:34 AM
Kind of weird that migrated questions like this one accumulate upvotes but the user never gets the benefit.
1 hour later…
10:00 AM
@yoda I guess that would make up for stopple choosing Vitaliy's answer over mine.
10:23 AM
Why would somebody copy an old question of my word by word and re-post it?
10:51 AM
@JM Hi, one more delete vote please? mathematica.stackexchange.com/questions/1488/…
@Szabolcs I don't have deletion capabilities yet... give me a few days or upvotes, whichever counts. :)
@Verbeia S/he will if s/he registers here with the same e-mail address used in the source SE site...
@Szabolcs I thought having 3000 rep would allow me to cast delete votes on questions, but I can't for that question for some reason.
@Heike strange... I know a few days must pass before one can cast the first delete vote unless having 4k rep, but MrW was able to case the second delete vote even though he is below 4k
11:11 AM
@Szabolcs Maybe it's because I voted to close on that question
Apparently one of the privileges of trusted users is "Voting to delete questions immediately after they are closed (even if they are new)", so I guess having less than 4000 rep only allows me to either cast a close or a delete vote on new questions
Ah, got it: "You must wait for a question to be closed for 2 days before you can vote for deletion"
Hi all. It was me who cast the second delete vote, not Mr.Wizard
That certainly demystifies things...
@JM I updated the answer again with a custom table function
11:29 AM
@Szabolcs regarding your custom table function: all good, except it is not re-entrant. This may be a problem, and finding this issue would be very hard
@Leonid ohhh ... you are right ...
@Szabolcs To avoid this problem, you need some instantiation mechanism, like making an obkect instance
will fix later, must leave for a while
will remove it til then
@Szabolcs Or you could just state the current limitation. To be more general, you could also take my function from here: stackoverflow.com/questions/6470625/mathematica-table-function/…, and modify that one in a similar manner (the final version of it is at the bottom).
@Szabolcs One more thing is that generally a List head is not sufficient to be used for linked lists, since the elements themselves can be lists. I am sure you know this, I just thought it is good to note, in case you want your function to be general.
@Leonid that's why I wrapped everything in a different head
If I had taken the time to look up Flatten's 3rd argument, I could have used that...
11:42 AM
@Szabolcs Oops, you did. Sorry - I should be reading code more carefully.
@Szabolcs yes, this is pretty much the standard way to do it (third argument of Flatten), AFAIK - this is perhaps why I did not "pattern-match" your code
11:54 AM
@All got to go. See you later.
@Leonid Also, it is re-entrant!!!! But only the inner table results can be monitored. I vow not to open this site before tomorrow morning, it is seriously damaging my short term real-work productivity and my todo list is already way too long for today ... :P (I think I'm addicted in a bad way)
12:17 PM
Hi @DavidCarraher. Did you see the last comment I addressed to you?
@Szabolcs looks like you are right again, although I'd have to give it more thought. I seem to be out of shape today. Anyways, must leave now, also have tons of stuff to do.
Hi @J. M. Yes, I take it that your code was sometimes returning the reflection of the inscribed triangle you were hoping to show, right?
@Heike Don't delete your answer, I actually wanted to mention PrintTemporary, I think it's good to have it here
@Heike I really don't think there's a way to see what an already running Table does ... so all we can do is figure out ways to monitor table for the next time a long calculation is launched
12:35 PM
@Szabolcs OK, I'll undelete it.
I'll try to beef it up a bit later, but I have to get some lunch first.
@DavidCarraher Yes. The angles sometimes came out wrong. The new version hopefully does not suffer from that bug.
2 hours later…
3:07 PM
Should this question be deleted? Please see the comments (the OP is asking for deletion)
Can you take a look at how I retagged this and correct if necessary?
@yoda if is not appropriate there, then where do we use it? Do we use it at all?
@Szabolcs I'm not sure. It could serve as an example of some of the unforeseen side effects that could pop up if you start redefining built in functions, but maybe it's better if he adds a warning to his StringJoin question.
3:31 PM
@yoda okay, since I posted that most -tags got removed
@Szabolcs he can delete it himself... no votes, no answers. he simply needs to click delete
@Szabolcs Our edits overlapped... I checked to see what you had done and at a glance, I only saw the title was edited and the thanks was removed, so I overrode it
I'm not sure about ... I don't feel strongly either way (with a slight leaning towards not having it), so you can add it back if you'd like
@JM did you change your avatar again?
The matrix square root question? Maybe there's a better additional tag than ...
@Heike Yup.
(I have way too much unreleased art in my hard drives...)
maybe you can set up a script that automatically rotates avatars
Something that works nicely with Gravatar? I'm not sure how to hijack their system for that...
3:46 PM
Mod[12,0.1] gives 0.1 despite 120*0.1 == 12. Mod[12, 0.25] gives 0. I know it's because of the binary representation of floating point numbers, but is there an easier workaround than Rationalize?
Note that == doesn't suffer from this.
Mod[12, 1/10] seems okay, as well as 12 - 0.1 Quotient[12, 0.1]. Something weird with Mod[]...
QuotientRemainder[12, 0.1] also works as expected.
@JM well, 0.1 is really less than 1/10 because 0.1 is not representable exactly in binary, so it gets slightly changed
But Equal tries to be smart about these things and 1/10 == 0.1 is True, so I'd expect Mod to be smart too
@Szabolcs I know that; note however that QuotientRemainder[] behaves nicely here...
Hm, funny ... I'd call it a bug
...so it's something peculiar to Mod[].
3:52 PM
@JM QuotientRemainder fails for QuotientRemainder[12.1, .1]
Something interesting: Table[Mod[12, N[1/10, k]], {k, 5, 25}]
So does Quotient[12.1, .1] for that matter
I wouldn't call it a bug
The documentation clearly says it's an integer mathematical function
although, later on it says it can be any numeric quantity, not necessarily integer
@yoda I know, I know... but there are applications that I have to (ab)use it, like Mod[10 Pi/3, 2 Pi].
I would've taken that to mean 1. is allowed, and not necesasrily 1 only
3:56 PM
Probably the whole point is that Quotient[]/Mod[] are unreliable for inexact numbers.
and now we can't assume any more that Quotient[a,b] * b + Mod[a,b] == a
I didn't mind that Mod[12,0.1] gave 0.1, but I do mind that it is inconsistent with Quotient
Something like Quotient[12.1/.1, 1] or .1 Mod[12.1/.1,1] seems to work (for this particular example at least)
On that note: Mod[x, 1] == SawtoothWave[x].
machine learning beta is not doing great... 1 day, 12 questions, 3 closed, 0 accepted answers
coming to think of it, dsp wasn't that great either... I guess mma private beta has raised my expectations too high
@yoda Maybe the machines are taking over on that beta
4:09 PM
@yoda That went into beta? I thought CV was supposed to welcome machine learning questions...
On the other hand, I would hate to see the DSP site go...
...it's not quite math; it's not quite engineering; and it's definitely not programming.
How long does a proposal need to be in beta before they start giving prognoses?
I don't know... lit.SE has been 180 or so days in public beta, for instance...
@JM That was hashed out during the area 51 process and the ML community wanted to be separate, because CV would look at it only from a pure statistics view point
@JM I don't think the dsp site is going to be culled... it's growing slowly (very slowly), but a dsp course every semester somewhere in the world brings us traffic... and questions (some low, but they get nuked instantly)
Still, one would hope to attract more professionals...
@Heike after 30 days, they introduce shame colouring on the A51 page telling us where we're slacking... disregard that. After that, they say they take a hard look at the site and think of extending (default) or graduating, but no way it's happening right after 90 days... it used to be the case when they first started launching sites, but with about 80ish sites and more being launched regularly, it's getting hard on them
@JM we actually did have a lot of professionals... a very good number. A lot of them left because the site stagnated because SE wouldn't open it to the public because of low volume and we didn't have the ability to invite others by email back then.
You can't put 200 monkeys in a closed room and force them to ask questions without giving them bananas... eventually they all started sulking and bickering
4:20 PM
"because the site stagnated because SE wouldn't open it to the public because of low volume" - ah, feedback loops... tsk.
@yoda I hope they'll let us graduate eventually. As far as I can tell we're doing quite well compared to other proposals in the beta phase.
Emphasis on "quite". This site's rates are fantastic!
@Heike I think they will... it'd be a shame if they left it in beta for a year like some sites
Lit.SE I think is the current longest-running public beta now.
whaat? no. There've been sites for nearly 2 yrs
4:26 PM
@JM What about personal finance and money?
I'm pretty sure gardening came before lit. Then there's onstartups, personal finance, rpg, bcg, etc
Ah, right. Gardening...
@yoda Gardening: 244 days
User Experience was in beta for something like a year + a few months
4:27 PM
Literature: 182 days according to area51
Holy feck, a year and a half?!
I believe they've dangled the graduation fruit a little too many times at rpg... if I were a member there from the beginning, I would've been mighty disappointed and pissed
First, teasing them with a design, then promises of graduating "soon" and the "numbers looking solid"...
@JM are you a fan of father Ted perhaps?
@Heike Well, British humor in general...
@JM The use of the word feck made me wonder
4:33 PM
@JM to be fair, most sites have had some problem or the other... take this for example.. good traffic, decent accept % and # of answers, low q/day, but no avid users
we have 3 users with more rep than their top user and more users at every rep level than they have
@TimStone they graduated very recently, right?
I believe just over a month ago
User Experience had 6.7 questions per day at the end of their beta.
but they did have almost 9000 users.
Ah, that's the problem... it wasn't OVER 9000!!!!1!
they migrated over from an SE1.0 site which was already active
(as did onstartups, money.se, EE and a few others)
How did SE1.0 work? Did you have to go through a beta phase as well?
4:42 PM
I do know for a fact that even SO in the olden times was in beta...
@Heike I think people had to pay for SE1.0, not sure though
Posted by Joel Spolsky on April 13th, 2010

Like the small-town mayor who suddenly finds herself running an entire state, our ambitions for Stack Overflow keep growing. Our original idea of making the Internet a better place to get expert answers to your programming questions suddenly seemed too small. Programming questions? We asked. Why just programming questions? Why not every question under the sun? And who says we can’t run for Vice President of the United States of America?

We tried making our software available as a hosted white label product called Stack Exchange. We thought that other people would create awesome sites on every i …

Did anyone here have problems with aborting parallel calculations before?
Not here
In command line mode first an abort does not work (ctrl-c, then a). The second time I press ctrl-c it says "Interrupt during LinkRead". Then aborting works and all is fine (parallel kernels get aborted as they should). But when I'm doing the same with the FE and try to abort with alt-., it sometimes just won't abort
A simple ParallelTable[Pause[1]; i, {i, 10}] can be aborted just fine though ...
@Szabolcs Wow, apparently it used to cost $129-$5000/month to have a site on stackexchange
@Szabolcs I find I usually need to whale on Ctrl+. until if finally aborts
4:50 PM
I just tried with something not pausy and worked fine
1000] //. {b___, i_, j_, a___} /; Abs[i - j] < 0.2 :> {b, i,
j + 0.2, a}, {100}]
I just tried with something not pausy and worked fine
That works for me too
I have to examine my real code more closely (it's a big monster with LibraryLink extensions, but I made sure that my LibraryLink code is interruptible)
5:09 PM
What I'd really like to have is a keyboard shortcut to wrap the selection in a function. It would surround everything with [ ] and put a Placeholder[f] in front
Is there a good way to highlight all initialization cells (even better: initialization groups)? I need to know which one is and which one isn't an init cell, because all this gets exported to a "package file" for running remotely
What I'd like is to get to 1000 rep now. I'm attentive to new questions, hehe
@Rojo You're nearly there...
@EliLansey He is there
Funny how rep can motivate people! :D
@Heike Just now. Congrats!
5:22 PM
:) ;:):) :)
ni this important moment of my life
Thank you all for being with me
@Szabolcs, you could post a question about highlighting init cells. Would be useful
@Szabolcs Not a shortcut, but you could do something like this
RowBox[{"\[Placeholder]", "[", "\[SelectionPlaceholder]", "]"}],
5:39 PM
@Heike thank you, works great!
5:53 PM
I must always wrap everything in MemoryConstrained .... I just took down a server with a small mistake, and I don't have the key to the server room to reboot ...
@Szabolcs Good move!
@Szabolcs Sounds like a good idea, if you're always doing the things that bring servers to their knees...
@Szabolcs Hmm, that sucks. At least your servers are near by: mine are 1500 miles away.
Is there a way to make any message abort immediately?
Check still evaluates the first argument fully, I think
@Szabolcs I accidentally wrote Nest[{#, #} &, x, 40] instead of Nest[{#, #} &, x, 4] on my laptop to create a binary tree... not fun. Took the whole thing down
5:58 PM
@Szabolcs any tips on running mma on a remote server? i've never been able to get parallel calculations to work on remote nodes
I hate swapfiles and swapping ...
well, at least, I could restart it, yes
@EliLansey it was a lot of trouble, but I managed, I'll tell you another time how
@All, my timeline for finishing got a little cramped today, so I'm withdrawing from the potential pro tempore moderator list.
"bringing servers to knees", "swapping" dunno, a lot of this sound pretty dirty to me
6:00 PM
@EliLansey Off the gutter, you. :P
@EliLansey using standard Unix commands as your phrases, you can write something that is rather explicit. And, no, I won't repeat it here.
@rcollyer i think i've seen something like that before
@rcollyer The file system check utility is useful in that regard. Sometimes the unarchiver for ZIP files as well.
@EliLansey yep, all physicists I know are perverts.
@JM yes, more, and finger ... and I'll leave it at that.
6:03 PM
mount and unmount as well
...and touch
Now that the G rating has run screaming, and the PG rating won't go near us, I think I'll call it a day.
@rcollyer sorry....
I once had to site through a seminar on Microsopy where they were talking about techniques like FRAP-ping, FLIP-ping, FLAP-ping and FRET-ting.
@Heike here in NY they're having discussions about legalizing frakking
6:06 PM
I didn't manage to keep a straight face
@EliLansey don't be sorry, I helped.
@Heike I don't quite believe you were expected to keep one... ;)
@EliLansey yes, they are ...
@rcollyer it's always hard not to chuckle when you hear the NPR reporters talking about frakking
Especially when you think of them as nice staid reporters.
6:09 PM
@rcollyer precisely
@EliLansey what year are you in?
@rcollyer lost track :-) hopefully my last...
it helps to have a dirty mind... I was taken aback at first when someone asked me how to "do a cronjob"
@EliLansey you too! I have to finish this year, the grad school kicks me out at the end.
What is frakking?
6:10 PM
I found it ... it was a bloody floating point number in Take!!
@rcollyer they're not going to kick me out, but my wife might kill me if I stay here longer
@Heike conjoined word: hydraulic fracturing.
Hydraulic fracturing is the propagation of fractures in a rock layer caused by the presence of a pressurized fluid. Hydraulic fractures form naturally, as in the case of veins or dikes, and is one means by which gas and petroleum from source rocks may migrate to reservoir rocks. However oil and gas companies may attempt to accelerate this process in order to release petroleum, natural gas, coal seam gas, or other substances for extraction, where the technique is often called fracking or hydrofracking. This type of fracturing, known colloquially as a frack job (or frac job), is done from ...
@EliLansey I'm being threatened daily, so I empathize.
ah, ok
6:11 PM
aka the cause of earthquakes in the UK
@rcollyer where/what are you studying?
and the fluid is not necessarily safe
@EliLansey LSU
At least that's where they say. I'm living in upstate NY now.
@rcollyer i was wondering why LSU would be talking about upstate fracking...
@rcollyer i'm at CUNY
@EliLansey as you said in your profile.
Eli, field?
@rcollyer ah yes, forgot about the whole profile thingy...
@rcollyer photonic metamaterials
6:14 PM
theory or experiment?
or computation?
@rcollyer mostly theory and computation. they keep trying to get me to do experiments, but I keep explaining how the equipment is too expensive for me to work with
@EliLansey Eep, those things are rather delicate, no?
@EliLansey the experimentalists at LSU make the same jokes about the theorists. In some cases, they're right.
But, give me a computer to rebuild, and I'm your man.
@JM yah. fragile little buggers. fortunately we have a team up at cornell CNF doing the really delicate stuff, and I only get their finished products to break
@rcollyer i hear that. for a while the joke in the lab here was that i was doing a phd in IT rather than physics
@EliLansey That's good!
But, in my case, one of the experimentalists had to come to me for access to the cluster.
I secured it for my adviser, and put it into (semi-)working condition.
anyway, I've got to go write something on my dissertation today. Talk to you later.
6:18 PM
@rcollyer ok, good luck!
Q: How to abort on any message generated?

SzabolcsIs there a way to immediately abort when any message is generated? This abort should work when doing parallel calculations as well. Ideally the first message would still be printed. I just brought a machine down because its memory got filled up due to a minor mistake.

1 hour later…
7:31 PM
@Eli If you tried to use the cluster integration package: it is terrible, and it doesn't give any error messages, it just fails ... mail me some details about what sort of machine you were trying to run Mathematica on remotely, and maybe I can help (if it is similar to our setup)
@Szabolcs yeah, i tried using their cluster integration thingy. thanks, i'll send you an email
@Eli I'll reply you sometime tomorrow, it's 10pm here, and I still have to walk home. Are you in NY? I wonder what it's like there, I heard the US is not very "walkable"
@Szabolcs Yeah, no rush at all. I'm in NYC, which is walkable / public transportable. But I live in NJ, so I tend to drive.
You know of some nice stylesheets to download?
7:58 PM
@halirutan, pity you can't do block-like highlighting
I guess I'll keep on using that "LetL" function of Leonid (that parses as nested With's that can depend on the previous one), with another syntax, lists of {lhs, rhs} instead of {lhs=rhs}, in order for it to highlight the locals
@Rojo have you read my comment?
Anyway to tell Mathematica to spit out factorials and not Gamma functions?
When you want to highlight the variables in red, if they are already defined in an other block, you have to be aware of this outer block and you have to parse it. I assume, they have in addition to SyntaxInformation an on-the-fly parser which parses small parts.
Yeah, makes sense
In any case, that's more than I would hope for. Probably it also means you also can't highlight like Table, but with {lhs=rhs} syntax
8:08 PM
I tried some combinations, but an expression like x=blub was never recognized.
Let's wait for the excuse of the support.
8:20 PM
Like... if the output has a Gamma[n] in it, anyway to tell Mathematica to translate these into (n-1)!?
8:37 PM
@LordStryker You could always do something like expr /. Gamma[a_] :> Factorial[a-1]
Hi all! @Rojo Just stopped by to say that when I am done with the Mathematica highlighting based on "Program" - style cells and my highlighter generator, you will be able to use any syntax you want, and highlight it in any way you want.
@LeonidShifrin looking forward to it :)
@Heike This is so close to what I need, however, I cannot figure out where to place it in the evaluated statement (I'm integrating). Is there a way to make a global statement regarding this?
@Rojo Just hope to get some time for that soon. There are just too many things I've got to do. Got to go now...
@Heike Actually a global def would be bad in my case. When I place it at the end of the integral it puts 'expr' into my output
8:45 PM
@LordStryker, perhaps something like Unprotect[Gamma];
Gamma[i : Except[_?NumericQ]] := (i - 1)!
@Leonid, bye bye, keep up the good work, hehe
@Heike Okay I removed 'expr' and it worked.
@LordStryker the expr was just a place holder for your actual expression
@Heike Yeah I realized that too late. DOH!
9:19 PM
All, so I do something like f[3] = 42;f[4] = 55; and I want to find out for which x f[x] has been defined.
I can do (First /@ DownValues[f]) /. Verbatim[HoldPattern][x___] -> x
can anybody see any simple way in which this will break?
(ie, am I doing something stupid?)
x__ will be the f[3] or f[4]
Make it Verbatim[HoldPattern][f[x___]]
that's true it gets evaluated this way
or HoldPattern[f[x___]]
in case that evaluates
ok thanks, that is a good point
what I'm after is, can you or someone else see some cases that will be missed?
Well, if you did f=g and then defined g[2], you won't see f[2] has a value
9:27 PM
@acl f[x_] := 33
good point. let me fix that
@EliLansey yes, also true (I am aiming for pattern-free downvalues though)
(but thanks)
But if you define something like that, the output is {HoldPattern[f[3]], HoldPattern[f[4]], HoldPattern[f[g]],
HoldPattern[f[h]], HoldPattern[f[x_]]}
rather than {3,4,g,h,x_}
or whatever that last term should be
@EliLansey, make it (First /@ DownValues[f]) /.
Verbatim[HoldPattern][HoldPattern[f[x___]]] :> x
@Rojo Bingo.
@EliLansey yes, what I mean is that I am applying this to the list after I filter out the ones with patterns, like so Select[DownValues[h],FreeQ[First[#], Pattern] &]
9:33 PM
@acl Ah, ok. In any case, @Rojo's last version picks those out, too
@acl, patterns can have head Blank, BlankSequence, BlankNullSequence
@Rojo but FreeQ[x___, Pattern] evaluates to False
or did I misunderstand you?
(ie try FullForm[x___])
Yeah, but without the x
FreeQ[___, Pattern]
or FreeQ[_, Pattern], or FreeQ[, Pattern]
All give True
oh, i see what you mean
good point
Sorry, I meant FreeQ[_
you're leaving definitions such as
f[x:{1, 2, 3}]:=8
just because they decided to name the argument
Probably makes more sense to filter Blank, BlankSequence and BlankNullSequence and leave Pattern be, I guess. Don't you think?
9:42 PM
hm probably
or maybe I should just forbid reuse of my code and forget about making it robust
Do you know of any site with some stylesheets apart from the default ones around?
My English gets weirder every day
@acl always a good technique :)
@Rojo I think there are some here home.comcast.net/~djmpark/Mathematica.html
David Park (he also has some discussion of how to create stylesheets). I'm not sure if they are nice or not though
10:03 PM
Hi all, I'm back in Sydney. Thought you might be interested in the traffic/user graphs.
@Verbeia I read somewhere, that SE or SO has now a final api for accessing such data. I assume you are using this. Say I want to have an overview about this, where would I start to read? Is your Mathematica-code for those graphics public?
@halirutan It's something I modified from something I think Szabolcs did. The key code is:
`getBetaPage[page_Integer] := Module[
ad = Import[
"http://api.stackexchange.com/2.0/users?page=" <>
IntegerString[page] <>
"&order=desc&sort=reputation&site=mathematica", "JSON"][[2,
{"display_name", "creation_date", "reputation",
"reputation_change_week", "is_employee", "last_access_date",
"user_type"} /. # & /@ ad
page\[Beta] = 0;
result\[Beta] =
FixedPoint[Join[#, getBetaPage[++page\[Beta]]] &, {}];,
StringForm["Loading page `` ...", page\[Beta]] ]`
The main trick is that the absolute time used for creation_date and last_access_date are not the same as Mathematica AbsoluteTime. You have to add 70 years.
10:13 PM
@halirutan this?
@acl yes, that was what I saw. Thx
@Verbeia, @acl really nice, thank you.
@halirutan thanks - it wasn't particularly original code. But it's nice to see how we are going, not daily but every so often. I should mention that the first graph of traffic and avid users is scraped by hand into a spreadsheet from the Area 51 page for the site.
10:50 PM
@LordStryker None that I know of, but one could always use /. and an appropriate rule...

« first day (21 days earlier)      last day (4400 days later) »